Primary microglia were cultured from the cerebral cortices of 1- to 3-day-old Sprague-Dawley rats as described previously (Pyo et al., 1999).Briefly, the cortices were triturated into single cells in minimal essential medium (MEM, Invitrogen, Carlsbad, CA) containing 10% fetal bovine serum (FBS, HyClone, Logan,UT) and plated in 75-cm2 T-flasks (0.5 hemisphere/flask) for 2-3 weeks. Microglia werethen detached from the flasks by mild shaking and filtered through a nylon mesh to remove astrocytes. Cells were plated in 6-well plates (7 × 104 cells/well), 60-mm dishes (5 × 105 cells/dish), or 100-mm dishes (106 cells/well). One hour later, the cells were washed to removeunattached cells before being used in experiments. BV2 immortalizedmurine microglia cells were from Dr. E. J. Choi. The BV2 cellline was grown in Dulbecco's modified Eagle's medium (DMEM, Invitrogen) and supplemented with 5% FBS. Cells were serum-starved overnightbefore treatment withgangliosides.
C. Electrophoretic Mobility Shift Assay (EMSA)
Cells were harvested and suspended in 9 times packaged cell volume of a hypotonic solution (10 mM HEPES, pH 7.9, 10 mM KCl,0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol (DTT), 0.5 mM phenylmethylsulfonylfluoride (PMSF)) including 0.5% Nonidet P-40. Cells were centrifugedat 500 × g for 10 min at 4 °C,
and the pellet (nuclear fraction)was saved. The nuclear fractions were resuspended in a buffercontaining 20 mM HEPES, pH 7.9, 20% glycerol, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, and 1 mM PMSF, incubated on ice for 60 min with occasional gentle shaking,and centrifuged at 12,000 × g for 20 min. The crude nuclear proteinsin the supernatant were collected and stored at 70 °C until used.
EMSA was performed for 30 min on ice in a volume of 20 µl, containing4 µg of nuclear protein extract in a reaction buffer containing8.5 mM EDTA, 8.5 mM EGTA, 8% glycerol, 0.1 mM ZnSO4, 50 µg/mlpoly (dI-dC), 1 mM DTT, 0.3 mg/ml bovine serum albumin,6 mM MgCl2, and γ-32P-radiolabeled oligonucleotide probe (3 × 104 cpm), with or without 20-50-fold excess unlabeled probe. In supershiftexperiments, protein extracts were incubated with 0.2-0.5 µg ofSTAT1 and STAT3 antibodies (Santa Cruz Biotechnology) for 30 minprior to the addition of 32P-labeled probe.
DNA-protein complexes were separated on 6% polyacrylamidegels in Tris/glycine buffer. The dried gels were exposed to x-rayfilm. The following double-stranded oligonucleotide was used inthese studies: GAS/ISRE, 5'-AAG TAC TTT CAG TTT CAT ATT ACT CTA-3', 27bp (Santa Cruz Biotechnology, sc-2537). 5'-end-labeled probeswere prepared with 40 µCi of [γ-32P] ATP using T4 polynucleotide kinase (Promega, Madison, WI) and were purifiedon Quick Spin Columns Sephadex G-25 (Roche MolecularBiochemicals, Mannheim, Germany).
D. Reverse transcriptase-polymerase chain reaction (RT-PCR)
Total RNA was extracted using RNAzol B (Tel-Test Inc., Friendswood, TX)
and cDNA was prepared using avian reverse transcriptase (Takara, Shiga, Japan), according to the manufacturer'sinstructions. PCR was performed with 30 cycles of sequential reactionsas follows: 94 °C for 60 s, 55 °C for 30 s, and 72 °C for 90 s.
Oligonucleotide primers were purchased from Bioneer (Seoul, Korea). The sequences of PCR primers were shown on Table 1. PCR products were separated by electrophoresis on a 1.5% agarose gel and detected under UV light.
E. Western Blot Analysis
Cells were washed twice with cold phosphate-buffered saline and then lysed in ice-cold modified RIPA buffer (50 mM Tris-HCl, pH 7.4, 1% Nonidet P-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM Na3VO4, and 1 mM NaF) containing protease inhibitors (2 mM PMSF, 100 µg/ml leupeptin, 10 µg/ml pepstatin, 1 µg/ml aprotinin, and 2 mM EDTA). The lysates were centrifuged for 10 min at 12,000 × g at 4 °C, and the supernatant was collected. Proteins were separated by SDS-PAGE and transferred to a nitrocellulose membrane. The membrane was incubated with primary antibodies and peroxidase-conjugated secondary antibodies (Vector Laboratories, Burlingame, CA) and then visualized using an enhanced chemiluminescencesystem(Sigma-Aldrich).
F. Determination of NO Release
Media nitrite concentration was measured as an indication of NO release.
Following the indicated cell incubations, 50 µl ofculture medium was removed and
mixed with an equal volume of Griessreagent (0.1% naphthylethylenediamine, 1%
sulfanilamide, 2.5%H3PO4), and absorbance of the mixture at 540 nm wasmeasured.
G. Enzymatic Digestion of Sialic Acid
Neuraminidase derived from A. ureafaciens was used for cleaving sialic acids residues from gangliosides. Gangliosides were dissolved in 10 mM sodium acetate buffer, pH 5.0, containing 1µg of sodium cholate per µl and were incubated with A. ureafaciensneuraminidase (Sigma-Aldrich) at 37 °C for 2h.
H. Oxidation of LDL
I was prepared oxLDL using a standard method of 2,2’-azobis(2-amidimopropane) dihydrochloride (AAPH; Sigma-Aldrich)-mediated oxidation (Neuzil et al., 1998; Shie et al., 2004). Oxidation of LDL was performed at 37 °C under the following conditions to obtain different degrees of oxidation. First, LDL was oxidized by 10 mM AAPH for 2, 6, or 18 h. Second, oxidation of LDL was performed for 18 h at AAPH concentrations of 1, 5, or 10 mM. Produced oxLDLs were represented as “oxLDL-concentration of AAPH (mM): oxidation period (h)”.
I. Relative electrophoretic mobility (REM) assay
Electrophoretic mobility relative to LDL was measured by agarose gel (0.8%
agarose in 0.08 mol/L Tris-HCl buffer, pH 8.3) electrophoresis and Coomassie Brilliant Blue R250 staining. This allows detection of changes in electric charge
induced by oxidation (Napoli et al., 1999; Sparks and Phillips, 1992).
J. Thiobarbituric acid reacting substances (TBARS) assay
200 µl of LDL/oxLDL (100 µg) was added to a test tube containing 200 µl SDS (8%, w/v), 400 µl acetic acid (20%, w/v), and 400 µl of thiobarbituric acid (0.8%, w/v). The mixture was vortexed well and boiled for 1 h. After cooling, the specimens were centrifuged (13,000 rpm, 10 min) and the absorbance of the supernatant was determined at 540 nm using a spectrophotometer. The amount of TBARS was determined by comparison to a standard of malondialdehyde (MDA) equivalents prepared using 1,1,3,3-tetraethoypropane (Sigma-Aldrich).
K. Lipid hydroperoxide (LPO) assay
The LPO level in oxLDL was determined by a LPO assay kit provided by Cayman. 100 µg LDL/oxLDL were used for LPO measurements. The absorbance at 500 nm was measured using a spectrophotometer (Amersham Pharmacia Biotech, San Francisco, CA).
L. Enzyme-linked immunosorbent assay (ELISA)
TNFα and MCP-1 levels in cell culture media were determined by ELISA as described by the manufacturer (OptEIA Sets, Pharmingen, San Diego, CA). TNFα and MCP-1 concentrations in the media were determined by spectrophotometer and calibrated from standards containing known concentrations of the cytokines.
M. Data analysis
Data were expressed as mean ± S.E.M. Analysis of variance followed by Dunnett’s multiple comparison tests were used for statistical comparisons.
III. RESULTS
A. JAK-STAT Signaling Mediates Gangliosides-induced Inflammatory Responses in Brain Microglial Cells
1. Gangliosides Induce Nuclear Factor Binding to GAS/ISRE Elements
Functional GAS/ISRE elements are found in the promoter regions of several inflammation-related genes, such as iNOS, and theseelements are known to bind the phosphorylated STAT dimer. In an attempt to explore the molecular mechanism of gangliosides on microglial activation, I investigated whether STATs could be involved in gangliosides-induced activation of microglia. I first examined the transcript level of iNOS in gangliosides-treatedrat primary microglia. Gangliosides markedly induced iNOS mRNAwithin 1 h (Fig. 1A, a), suggesting that gangliosides directly regulate NOproduction at the level of transcription. This observation was subsequently evaluated by EMSA using a γ-32P-labeled consensus GAS/ISRE oligonucleotides probe. After the cells were treated with 50 µ g/ml brain-derived gangliosides mixturefor the indicated times, nuclear extracts were prepared and then analyzed by EMSA. The specific binding complex was detected innuclear extracts from gangliosides-treated rat primary and murineBV2 microglia (Fig. 1, A(b) and B).
Time course analysis showed thatgangliosides rapidly induced the nuclear factor binding within5 min and that the binding activity was decreased to basal levelsafter
Fig. 1. Gangliosides stimulate iNOS transcription and nuclear factor binding to GAS/ISRE elements in microglial cells. (A) Rat primary microglial cells were treated with or without 50 µg/ml brain gangliosides mixture (Gmix) for 1 h, after which total RNA was isolated, and levels of iNOS mRNA were measured using an RT-PCR-based assay. The transcription of GAPDH was measured for normalization (a). Cells were treated with Gmix for 5 min, after which nuclear extracts were prepared and assayed for the amount of binding activity to GAS/ISRE oligonucleotides using EMSA (b). (B) BV2 cells were treated with 50 µg/ml Gmix for the indicated periods. Nuclear extracts were prepared, and binding activity to GAS/ISRE oligonucleotides was determined by EMSA.
Fig. 2. Gel shift assay using anti-STAT1 and anti-STAT3 in BV2 microglial cells. After cells were treated with 50 µ g/ml Gmix for 5 min, gel shift assays were performed as described in Fig. 1. with the exception that nuclear extracts were incubated with 0.2-0.5 µ g of STAT1(A) and STAT3 (B) antibodies for 30 min prior to the addition of 32P-labeled probe. *, inset is a short-exposed autograph of the upper band (dotted box).
30 min in both microglial cell types (Fig. 1B). The specificity of the shifted bands was confirmed by competition assay using excess amounts of unlabeled oligonucleotides(cold oligo). In addition, gel shift assay showed that the binding complex was diminished by addition of anti-STAT1 and anti-STAT3,indicating that both STAT1 and STAT3 are constituents of the nuclearfactor binding complex (Fig.
2). These results show that functional GAS/ISRE elements may be involved in gangliosides-induced activationof microglia.
2. Gangliosides Induce the Phosphorylation of STAT1 and STAT3
Essential roles for STAT signaling in brain inflammatory response have emerged. Because gangliosides rapidly induced the GAS/ISRE-nuclear factor binding, I examined whethergangliosides indeed caused phosphorylation of STAT proteins. Primarymicroglial cells were stimulated with 50 µg/ml gangliosides forthe indicated times, and the levels of phosphorylated STAT1 were determined by Western blot analysis using antibodies against Tyr-701-STAT1and Ser-727-STAT1.
Both phosphorylations of STAT1 occurred within5 min of gangliosides addition (Fig.
3A, a). Similar patterns of phosphorylation were observed in lysatesfrom murine BV2 microglial cells, where incubation of cells withgangliosides resulted in STAT1 phosphorylation on tyrosine and serine residues, with phosphorylation levels returning to basalat 30 min (Fig. 3B). In addition to phosphorylation of STAT1,I detected gangliosides-induced phosphorylation of STAT3 in both microglial cell types. The pattern of STAT3 tyrosine phosphorylationappeared similar to that of
Fig. 3. Gangliosides induce the phosphorylation of STAT1 and STAT3 in microglial cells. Rat primary microglial cells (A) and mouse BV2 microglial cells (B) were serum-starved for 12 h and then stimulated with 50 µg/ml Gmix for the indicated times. Cell lysates were separated by 10% SDS-PAGE and Western blots probed with anti-pSTAT1 (Y), anti-pSTAT1 (S), or pSTAT3 (Y). The membrane was then stripped and analyzed with anti-STAT1 antibody to determine loading.
STAT1 phosphorylation (Fig. 3, A andB). The Western blotting data show that gangliosides trigger rapidphosphorylation of STAT1 and STAT3, suggesting their involvement in gangliosides-induced microglial activation. The phosphorylation patterns of both STAT1 and STAT3 determined by Western blottingcorrelate with the binding activity results from EMSA.
3. Gangliosides Induce Phosphorylation and Activation of JAK1 and JAK2 Phosphorylation of STATs depends on the activation of JAKs. JAKs both functionally and physically associate with cytokinesignaling. In particular, activation of JAK1 and JAK2 providesa molecular explanation for cellular actions of a broad rangeof cytokines. Thus, I investigated whether JAK1and JAK2 could be involved in gangliosides-induced STAT phosphorylation. Primary rat microglial cells were stimulated with 50 µg/ml gangliosidesfor the indicated times, and cell lysates were Western blottedusing antibodies directed against phosphorylated JAK1 and JAK2.
The data presented in Fig. 4 show that following addition ofgangliosides to cells, phosphorylation of both JAK1 and JAK2 occurred within 5 min, after which phosphorylation levels returned to basallevels by 30 min. The involvement of JAK signaling in gangliosides-induced microglial activation was also shown using a second, independent approach. The pharmacological agent AG490 is known to inhibitthe phosphorylation of both JAK1 and JAK2. I found thatpretreatment of rat primary microglial cells with AG490 effectively reduced gangliosides-induced phosphorylation of STAT1 and STAT3(Fig. 5A). In addition, AG490 inhibited the
Fig. 4. Gangliosides stimulate phosphorylation of JAK1 and JAK2 in rat primary microglial cells. Cells were serum-starved for 12 h and then stimulated with 50 µg/ml Gmix for 5 min. The phosphorylation of JAK1 and JAK2 was determined by Western blot analysis using antibodies specific for phosphoJAK1 or -JAK2.
Fig. 5. AG490, a specific inhibitor of JAK, reduce gangliosides-induced STAT phosphorylation and nuclear factor binding to GAS/ISRE. (A) Cells were pretreated with 10 µM AG490 for 1 h and then stimulated with 50 µg/ml Gmix for 2 min. Western blots were probed with anti-pSTAT1 (Tyr-701) and pSTAT3 (Tyr-705).
The membrane was subsequently stripped and probed with anti-STAT1 and STAT3 antibodies. (B) Cells were pretreated with 10 µM AG490 for 1 h and then stimulated with 50 µg/ml Gmix for 5 min. Nuclear extracts were prepared, and binding activity to GAS/ISRE oligonucleotides was determined by EMSA. *, inset is a short-exposed autograph of the upper band (dotted box).
nuclear factor bindingto GAS/ISRE nucleotides in gangliosides-treated microglial cells(Fig. 5B). These results indicate that gangliosides induce phosphorylationand activation of STAT1 and STAT3 through phosphorylation andactivation of JAK1 and JAK2.
4. Gangliosides Stimulate STAT-responsive Inflammatory Gene Expression Brain inflammatory responses are coordinated by the production of cytokines, chemokines, and reactive oxygen species. The above data indicate that gangliosides-induced microglial activationmay be mediated, at least in part, by JAK-STAT-dependent transcriptionalresponses. Therefore, I examined the transcript level of genesthat have been reported previously to have functionalGAS elements and act as mediators of inflammation, namely MCP-1 and ICAM-1. Rat primary microglial cells and BV2 cells were stimulatedwith 50 µg/ml gangliosides for 3 h, and total RNA was extractedfor RT-PCR analysis. Addition of gangliosides rapidly increased the mRNA levels of both MCP-1 and ICAM-1, as did IFN-γ, whichwas included as a positive control (Fig. 6A). Pretreatment with AG490 significantly inhibited gangliosides-induced transcription of both genes (Fig. 6B). These findings demonstrate that gangliosidestrigger STAT-dependent transcriptional activation of inflammatorygenes in microglia.
5. AG490 Reduces Gangliosides-induced Release of NO
NO is known as an important physiological signaling molecule in the brain.
Fig. 6. Gangliosides stimulate transcription of STAT-responsive inflammatory genes in microglial cells and AG490 suppresses this transcription. (A) Cells were treated for 3 h with 50 µg/ml Gmix or 10 U/ml IFN-γ. Total RNA was isolated and analyzed for levels of MCP-1 and ICAM-1 mRNA using an RT-PCR-based assay.
The transcription of GAPDH was measured for normalization. (B) Cells were pretreated with 10 µM AG490 for 1 h and then stimulated with 50 µg/ml Gmix for 3 h. mRNA expression of ICAM-1 and MCP-1 was detected using an RT-PCR-based assay.
Aberrant iNOS expression and excessive NO production are observed in various pathophysiological conditions. Previously, I showed that gangliosides-induced microglialactivation was accompanied by induction of NO release. Thus, Itested whether gangliosides induced NO release via JAK-STAT signaling.First, I examined the effect of JAK inhibition on gangliosides-inducedtranscription of iNOS in rat primary microglial cells. RT-PCRanalysis showed that the inhibitor AG490 reduced mRNA levels ofiNOS (Fig. 7A). Second, I investigated the effect of AG490 onNO release. In these studies, the ERK inhibitor, PD98059, wasalso used since I have shown previously that it reducedgangliosides-induced NO release (Pyo et al., 1999).
In the presence of AG490, microglialcells were treated with 50 µg/ml gangliosides for 48 h, and theamount of NO produced was determined by measuring the amount of nitrite converted from NO in the media. AG490 significantly reduced gangliosides-enhanced NO release, as did PD98059 (Fig. 7B). Comparedwith cells treated with gangliosides alone, NO release was reduced to 38.6 ± 4.3 and 25.2 ± 14% in cells co-treated with PD98059and AG490, respectively. These results are consistent with theresults shown in Figs. 5 and 6. The findings indicate that JAK-STAT signaling is required for NO release and provide evidence of the critical functional involvement of JAK-STAT signaling in gangliosides-inducedmicroglial activation.
6. ERK Activity Appears to be Regulated by JAK Activation
There are several reports showing that the transcriptional activity of STATs
Fig. 7. AG490 reduces gangliosides-stimulated NO in rat primary microglial cells. (A) Cells were pretreated with 10 µM AG490 for 1 h and then stimulated with 50 µg/ml Gmix for 3 h. mRNA expression of iNOS was detected by RT-PCR analysis. The transcription of GAPDH was measured for normalization. (B) Cells were treated with 50 µ g/ml Gmix for 48 h in the presence or absence of AG490 or PD98059. The amount of NO was determined by measuring the amount of nitrite in the media, as described under "Experimental Procedures.".
is regulated through mitogen-activated protein kinases (MAPKs). MAPKs are considered as common intracellular signaling molecules involved in microglial activation. Previous reports by others and us showed that gangliosides induced activation of MAPKs in microglia. In the present study, I used pharmacological inhibitors to examine possible cross-talk between the JAK-STAT and MAPKs signaling pathways. When primaryrat microglial cells were pretreated for 2 h with the JAK inhibitor AG490, gangliosides-induced activation of ERK1/2 was significantlyreduced compared with controls with no AG490 (Fig. 8). In contrast,no significant suppression of p38 was observed under this condition.However, in the presence of PD98059, an ERK inhibitor, not onlyERK but also p38 activation was completely inhibited. These results indicate that gangliosides-stimulated JAK activation leads to activation of ERK in microglial cells. These pharmacological studiesalso indicate that gangliosides-stimulated activation of p38 maynot be due to activation of ERK by JAK.
7. Sialic Acid Residues Are Important for Gangliosides-induced Phosphorylation of STAT
The major types of gangliosides in brain are GM1, GD1a, GD1b, GT1b, and GQ1b. These gangliosides differ with respect to thenumber and position of sialic acid residues attached to the carbohydrates. The approximate percentages of each gangliosides presentin the brain gangliosides mixture used in the current study are 18% GM1, 55% GD1a, 15% GD1b, 10% GT1b, and 2% others. To addresswhether
Fig. 8. Activation of ERK1/2 follows JAK-STAT activation in gangliosides-treated primary microglial cells. Primary microglial cells were pregangliosides-treated with AG490 or PD98059 for 1 h and then treated with 50 µg/ml Gmix for 30 min. Cell lysates were separated by 10% SDS-PAGE and Western blots probed with anti-phospho-ERK and anti-phospho p38, respectively. The membrane was then stripped and probed with anti-ERK antibody. At least four experiments were independently performed, and representative data are shown in this figure.
the structural diversity of gangliosides affected activationof STAT, I compared the effect of GM1, which has one moleculeof sialic acid, with GD1a, which has two molecules of sialic acid, on phosphorylation of STAT1. Primary microglial cells were treated with GM1 or GD1a for 2 min, and levels of phosphorylated STAT1 were determined by Western blot analysis using antibodies againstTyr-701-STAT1.
The data in Fig. 9 show both GM1 and GD1a stimulatedphosphorylation of STAT1 within 2 min. The level of STAT1 phosphorylation stimulated by either GM1 or GD1a was similar to that caused bythe gangliosides mixture, suggesting that the number of sialic acid residues per gangliosides molecule has little effect on the phosphorylation of STAT1 in microglial cells (Fig. 9A). Becausesialic acid residues are characteristic of gangliosides, I examined whether sialic acid residues were important for gangliosides-stimulated STAT phosphorylation. Gangliosides were preincubated with either550 or 1000 units/ml A. ureafaciens neuraminidase, which is known to release sialic acid attached to an internal galactose in anygangliosides including GM1. Primary microglia cells were stimulated with gangliosides or neuraminidase-treated gangliosides(desialylated gangliosides) for 2 min, and levels of phosphorylated STAT1 were determined by Western blot analysis. The data presented in Fig. 9, B and C show a dose-dependent inhibitory effect of neuraminidasetreatment on phosphorylation of STAT1, indicating that sialic acid residues are required for stimulation of JAK-STAT signaling. To rule out the possibility that these reductions are due to contaminatingsialic acid or neuraminidase, I compared the effect of GM1 and asialo-GM1 (Sigma) on phosphorylation of
Fig. 9. The sialic acid of gangliosides is necessary for activation of JAK-STAT signaling. (A) Primary microglial cells were treated with 20 µ g/ml GM1 or GD1a for 2 min. Cell lysates were subjected to Western blot analysis, and levels of
Fig. 9. The sialic acid of gangliosides is necessary for activation of JAK-STAT signaling. (A) Primary microglial cells were treated with 20 µ g/ml GM1 or GD1a for 2 min. Cell lysates were subjected to Western blot analysis, and levels of